There's a phenomenon in scientific research that sometimes the act of performing or observing the experiment affects the outcome. To take a simple example, suppose a researcher wants to know the temperature of some liquid. So he inserts a thermometer. But the thermometer itself has some temperature, and so the act of inserting the thermometer may alter the temperature of the liquid. Or if you were trying to study whether people behave irrationally in some situation, if the subjects know you are observing them, they may be more careful in their behavior because they don't want to embarrass themselves in front of strangers. Etc.
Is there a word or term for this phenomenon?